Identifying general reaction conditions by bandit optimization.

Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries1-6. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models7-10 to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.

Reaction profiling by ultra high-pressure liquid chromatography/time-of-flight mass spectrometry in support of the synthesis of DNA-encoded libraries.

An ultra high-pressure liquid chromatography/mass spectrometry (UHPLC/MS) separation and analysis method has been devised for open access analysis of synthetic reactions used in the production of DNA-encoded chemical libraries. The aqueous mobile phase is 100mM hexafluoroisopropanol and 8.6mM triethylamine; the organic mobile phase is methanol. The UHPLC separation uses a C18 OST column (50mm×2.1mm×1.7µm) at 60°C, with a flow rate of 0.6mL/min. Gradient concentration is from 10 to 40% B in 1.0min, increasing to 95% B at 1.2min. Cycle time was about 5min. This method provides a detection limit of a 20-mer oligonucleotide by mass spectrometry of better than 1pmol on-column. Linear UV response for 20-mer extends from 2 to 200pmol/µL in concentration, same-day relative average deviations are less than 5% and bias (observed minus expected) is less than 10%. Deconvoluted mass spectra are generated for components in the predicted mass range using a maximum entropy algorithm. Mass accuracy of deconvoluted spectra is typically 20ppm or better for isotopomers of oligonucleotides up to 7000Da.

Novel inhibitors of the high-affinity L-proline transporter as potential therapeutic agents for the treatment of cognitive disorders.

The incidence of cognitive disorders such as Alzheimer's disease continues to increase unabated. While cures for such diseases have eluded investigators, progress is being made on alleviating certain symptoms of these diseases. Mouse knockouts of the proline transporter (PROT), a high affinity Na(+)/Cl(-)-dependent transporter, indicated its potential as a novel therapeutic target for cognition improvement. Herein we report our investigation into a novel class of PROT inhibitors.

Enantioenriched dihydropyrones from beta-lactone templates.

[reaction: see text] Optically active 4-substituted 2-oxetanones provide conduits for preparing 2,3-dihydro-4H-pyrone heterocycles. Enantioenriched beta-lactones are prepared by asymmetric catalytic acyl halide-aldehyde cyclocondensation reactions. Hydrazone anion-mediated beta-lactone ring opening and ensuing cyclization-dehydroamination of the derived beta-ketohydrazone afford the desired dihydropyrones (68-81%). Optimized lactone ring-opening-cyclization reaction conditions render a variety of optically active 4-substituted-2-oxetanones as effective precursors to enantioenriched 2,3-dihydro-4H-pyrones.